This proposal focuses on two types of change in synapse morphology: expansion of postsynaptic membranes and formation of presynaptic boutons. Anatomical specializations are a hallmark of synapses. On the postsynaptic side, anatomical specializations include dendritic spines and membrane folds. On the presynaptic side, an enlargement of the neurite to form a rounded presynaptic bouton or en passant swelling is a nearly universal feature of synapses. Aim 1 of this proposal tests a specific hypothesis for how activity can regulate anatomical changes at the postsynapse in the fly neuromuscular junction and in mammalian dendritic spines. In particular, it seeks to elucidate the role of Ral as a mediator of activity-dependent anatomical plasticity. We have uncovered a novel pathway for synaptic plasticity in which the small GTPase Ral, by activating the exocyst complex, serves a central role in transducing Ca2+ influx from glutamate receptors into enhanced transport of membrane to the postsynaptic region. In consequence, the membrane area of the postsynaptic specialization at the fly neuromuscular junction (NMJ) expands in an activity-dependent manner. This proposal asks what patterns of synaptic activity are necessary to recruit the exocyst and how Ral comes to be localized to postsynaptic membranes. It goes on to investigate the significance of Ral for the formation of dendritic spines in the mammalian CNS. Aim 2 uses Drosophila as a model system in which to uncover the machinery that allows synaptic boutons to form. The maturation of presynaptic terminals from growth cones to synaptic boutons is a critical late step in synaptogenesis but poorly understood. Neither the cytoskeletal elements that underlie the shape change nor signaling molecules that trigger it are known. In a previous mutant screen in Drosophila we discovered that mutations in 2-3 arrest synaptogenesis after initial synapse formation and prevent the formation of synaptic boutons at the fly NMJ. We propose to pursue the process of bouton formation in greater detail by a combination of biochemical and genetic approaches to uncover additional players in the formation of this poorly understood aspect of synaptic anatomy. Therefore, in Aim 2A we will use biochemical methods to identify binding partners for 2-3 and determine if they are required for bouton formation. In Aim 2B we will uncover additional players in bouton formation through a mutant screen of the Drosophila genome at the embryonic NMJ.